Perl 6 - the future is here, just unevenly distributed

IRC log for #marpa, 2014-05-23

| Channels | #marpa index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
01:21 jeffreykegler joined #marpa
05:56 jeffreykegler joined #marpa
05:57 jeffreykegler ronsavage: I remember that I promised you a few items about lexing -- things have come up and I'll need a few days.
05:57 jeffreykegler But I haven't forgotten!
07:04 ronsavage joined #marpa
07:21 ronsavage left #marpa
09:59 seki joined #marpa
10:00 seki hi
10:04 seki just a note to say that I have published my work-i-progress on marpa grammars to parse some PowerBuilder source files
10:05 seki see http://github.com/sebkirche/MarpaX-Languages-PowerBuilder for the grammars and semantics (work done in collaboration with my co-worker)
10:06 seki and http://github.com/sebkirche/PowerBuilder-DataWindow for a first diagnostic/analysis module built upon the previous
10:07 seki also, keep in mind that I am not a perl monk but only a beginner in Perl ;^)
10:07 seki so expect the code to might not be the state of the art...
10:08 seki for now I am quite satisfied with the semantics part and I am working on building tool with it
10:10 seki the PowerBuilder datawindow object is basically a view that can display data from a database (in free form display, grid, etc) similar to the MS Query tool or crystal reports
10:11 seki and it involve dealing with database queries
10:12 seki it can use 2 kinds of queries : a dedicated format to describe the query for which we have made a marpa parser
10:13 seki and the other kind is plain text sql query that I would like to parse also with marpa (if possible)
10:15 seki what could you advise me to do for that sql parsing part? I have found the SQL-AST from Jean-Damien Durand on http://github.com/jddurand/MarpaX-Languages-SQL-AST (untested yet) is it the way to go ?
10:18 seki for now my needs are only to get the table and field names from sql select statements, but the fields may use aliases or complex statements like casts, cases or just field concatenations so a true parser is needed to replace my poor man's field extraction via regexps as a first step
13:12 jeffreykegler joined #marpa
13:16 jeffreykegler jdurand: How is MarpaX::Languages::SQL:AST coming along?  Is it ready for seki?
16:12 seki left #marpa
18:29 jdurand_ joined #marpa
18:30 jdurand_ Re http://irclog.perlgeek.de/marpa/2014-05-23#i_8765231 - SQL grammar is ambiguous and I stopped when dealing with that
18:32 jdurand_ But it depends on what is parsed. t/create_schema.t is ambiguous. While t/select.t parsed successfully a very simple SELECT statement. In conclusion: not ready for production. That's why it has not been pushed to CPAN.
19:04 jeffreykegler joined #marpa
19:19 jdurand_ Jeffrey, Centos and RH bases distribs decided to revisit perl notion of core modules, their fault IMHO. Still adding Time::Piece will not hurt, until another appear

| Channels | #marpa index | Today | | Search | Google Search | Plain-Text | summary